HAL Id: hal-03136340https://hal.archives-ouvertes.fr/hal-03136340

Submitted on 9 Feb 2021

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Assessment of the effectiveness of wood pole repairusing FRP considering the impact of climate change on

decay and hurricane riskEric Merschman, Abdullahi Salman, Emilio Bastidas-Arteaga, Yue Li

To cite this version:Eric Merschman, Abdullahi Salman, Emilio Bastidas-Arteaga, Yue Li. Assessment of the ef-fectiveness of wood pole repair using FRP considering the impact of climate change on decayand hurricane risk. Advances in Climate Change Research, Elsevier, 2020, 11 (4), pp.332-348.�10.1016/j.accre.2020.10.001�. �hal-03136340�

Available online at www.sciencedirect.com

ScienceDirect

Advances in Climate Change Research 11 (2020) 332e348www.keaipublishing.com/en/journals/accr/

Assessment of the effectiveness of wood pole repair using FRP consideringthe impact of climate change on decay and hurricane risk

Eric MERSCHMANa, Abdullahi M. SALMANa,*, Emilio BASTIDAS-ARTEAGAb, Yue LIc

a Department of Civil & Environmental Engineering, The University of Alabama in Huntsville, Huntsville, AL 35899, USAb Universit�ede Nantes, GeM, Institute for Research in Civil and Mechanical Engineering, Nantes, CNRS UMR 6183, France

c Department of Civil Engineering, Case Western Reserve University, Cleveland, OH 44106, USA

Received 20 November 2019; revised 30 June 2020; accepted 14 October 2020

Available online 4 November 2020

Abstract

Electric power distribution systems are vulnerable to disruption due to severe weather events, especially hurricanes. Such vulnerability isexpected to increase over time due to the impact of climate change on hurricanes and the decay of wood poles that support the distribution lines.This study investigates the effectiveness of using fiber-reinforced polymer (FRP) sleeve to reinforce wood poles subjected to decay and hurricanehazard to restore their lost strength and extend their effective service life. The potential impact of climate change on the pole decay rate and theintensity and frequency of hurricanes is also considered. The optimal FRP repair time based on the structural reliability of the poles is alsodetermined. Three locations with varying climates are chosen to evaluate and compare the results: Miami, Charleston, and New York City. Theresults show that in all three locations, the application of the FRP sleeve can more than double the service life of the pole depending on the timeof the repair. The results also show that climate change significantly increases the vulnerability of the pole. The probability of failure of the poleat the end of the 21st century under RCP8.5 emission scenario in Miami, Charleston, and New York City increase by about 30%, 70%, and 73%,respectively, compared to a no climate change scenario. If climate change is only assumed to affect the decay of the pole, i.e., no change inhurricane hazard intensity, the corresponding increases in failure probability are 5%, 22%, and 20% in Miami, Charleston, and New York City,respectively. This implies that most of the impact of climate change on pole failure risk is due to the increase in hurricane intensity. The impactof climate change on decay is found to be comparatively small. It increases with time as variation in temperature and precipitation becomes moreprominent towards the end of the 21st century. The optimization results show that the optimal FRP repair time depends on how the FRP affectsthe pole's decay rate. If the FRP can significantly slow down the wood decay rate, the optimal time of repair is at the beginning of the pole's lifecycle. If the FRP has no impact on the wood decay rate, it is better to repair the pole after significant decay has occurred.

Keywords: Wood poles; FRP repair; Hurricane; Wood decay; Climate change adaptation

1. Introduction

Maintenance is a critical factor in the management of civilinfrastructure systems. A good maintenance schedule will savean owner substantially over the life span of a building orsystem, both economically and in terms of safety. For

* Corresponding author.

E-mail address: ams0098@uah.edu (SALMAN A.M.).

Peer review under responsibility of National Climate Center (China

Meteorological Administration).

https://doi.org/10.1016/j.accre.2020.10.001

1674-9278/Copyright © 2020, National Climate Center (China Meteorological A

This is an open access article under the CC BY-NC-ND license (http://creativec

electrical utility companies, the maintenance of electric powersystems prevents the loss of service and may push lower costsonto the end-user. As such, utility companies develop main-tenance plans for various components that are subjected towear and tear as well as deterioration due to natural causes. Anexample of such components is the poles supporting overheadnetworks. Hundreds of millions of poles that are worth billionsof dollars are used by utility companies across the U.S. andother countries (Mankowski et al., 2002; Ryan et al., 2014).

Wood poles are most commonly used to support electricaldistribution systems in the U.S. due to several factors. Wood

dministration). Production and hosting by Elsevier B.V. on behalf of KeAi.

ommons.org/licenses/by-nc-nd/4.0/).

333MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

is readily available in most places across the country or canbe acquired relatively cheaply. However, wood is an organicmaterial and is subject to decay and rot, which diminishes itsstrength over time. Consequently, the NESC (NationalElectric Safety Code) (IEEE, 2017) recommends that supportpoles for distribution lines be replaced once they reach two-thirds of their initial strength. Over time, deregulation anddownsizing of labor by utility companies make it moredifficult to maintain a reliable maintenance schedule, whichcan quickly identify failing poles (Mankowski et al., 2002).The rate of decay of wood is affected by climatic factors,especially temperature and moisture (Brischke and Rapp,2008; Teodorescu et al., 2017; Wang et al., 2008). It is,therefore, expected that potential future changes in temper-ature and precipitation due to climate change will have animpact on wood decay. However, such changes will onlyaffect the rate of decay, not the mechanisms causing thedecay. It can be said, therefore, that there is a correlationbetween climate change and the decay rate and subsequentfragility of wood poles.

Wood poles are also vulnerable to damage due to naturalhazards such as tropical cyclones, which are one of the mostcostly natural disasters in terms of both loss of property andlife (Gutowski et al., 2008). There is evidence that the in-tensity of tropical cyclones has been increasing over the years(Emanuel, 2005; Mann and Emanuel, 2006). The intensity isalso predicted to continue increasing in the coming years dueto climate change (Knutson et al., 2010; Mudd et al., 2014a,Oouchi et al., 2006; Staid et al., 2014). However, predictionson the potential impact of climate change on tropical cycloneshave a large degree of uncertainty because hurricane formationdepends on several climatic factors such as sea surface tem-perature (SST), North Atlantic Oscillation, Southern Oscilla-tion, El Ni~no effect, vertical wind shear, atmospheric stability,and other factors (Cui and Caracoglia, 2016; Ranson et al.,2014). As such, variability in hurricane activity is a complexconvolution of natural and anthropogenic factors. Variabilityalso differs based on the location under evaluation. Globalclimate models may over or underestimate changes in fre-quency and intensity of storms in individual basins (Emanuelet al., 2008; Knutson et al., 2010; Ryan et al., 2016).

To reduce the vulnerability of decayed wood poles to hur-ricanes and climate change, various strategies that involveregular maintenance, repair, or periodic replacement can beemployed. Ryan et al. (2016) proposed a framework to studythe impact of climate change on wood pole networksconsidering time-dependent decrease in strength due to decay.An advanced stochastic simulation method was proposed tostudy the cost-effectiveness of various climate change adap-tation measures for wood pole networks (Ryan and Stewart,2017). In the method of Ryan and Stewart (2017), variousadaptation measures based on alterations to design or main-tenance practices such as installation of larger poles and morefrequent inspections were considered. Another potentialstrategy is the use of fiber-reinforced polymers (FRP) to repairand strengthen decayed poles. FRP are strands of carbon-fiberreinforcement arranged in parallel and cast into an epoxy

coating and applied to the exterior of structural elements.Carbon-fiber has exceptional tensile capacity, so the applica-tion of FRP to the extreme tension fiber of a structural cross-section can increase the bending capacity of the structuralelement (Grace et al., 1999). The ease of application of FRPmakes it an ideal solution to reinforce existing structures. Forutility poles, FRP repair can be carried out by simply exca-vating around the pole and wrapping it with the FRP. As such,service interruption can be avoided (Grace et al., 1999).

Research on the use of FRP products to repair andstrengthen building and bridge components is abundant (e.g.,Sharif et al. (1994), Najm et al. (2007), Ghobarah and Said(2001), Triantafillou (1998), and Balsamo et al. (2005)).However, research on the effectiveness of using FRP to repairwood utility poles to improve their structural reliability isscarce. Saafi and Asa (2010) tested the moment capacity of 30-year-old decayed wood poles repaired using FRP. The authorsreported that the FRP could restore more than 85% of theinitial strength of the poles. Other research in which utilitypoles were repaired with FRP and tested to determineimprovement in strength include Polyzois and Kell (2007),Kell (2001), Chahrour and Soudki (2006), Chahrour andSoudki (2006), and Lopez-Anido et al. (2005). With the useof FRP for wood pole repair gaining popularity, it is essentialto investigate the long-term structural performance and life-span of FRP-repaired poles subjected to hurricanes. It is alsocrucial to investigate the potential impact of climate change onwood pole decay rate and the possibility of using FRP as aclimate change adaptation strategy to reduce the long-termvulnerability of wood poles.

This study aims to i) investigate the changes in thevulnerability of wood poles due to the impact of climatechange on wood decay and hurricane hazard, ii) investigate theeffectiveness of FRP repair on the long-term structural reli-ability of wood utility poles, iii) investigate the impact of thetime of FRP repair on the service life of the poles, and iv)determine the optimal FRP repair time to minimize thestructural reliability of the poles. Three coastal locations areconsidered to investigate the variation of the results fordifferent geographical conditions: Miami, Florida, Charleston,South Carolina, and New York City, New York. The results ofthe study will give an insight into the potential increase in therisk of failure of utility poles due to climate change. It willalso provide an insight into the effectiveness of using FRP torepair decayed poles to adapt to climate change.

2. Data and methods

2.1. Decay model description

Wooden poles are vulnerable to decay at its base where itcontacts the soil. The abundance of moisture and access tooxygen in the soil causes fungal growth and/or the presence ofinsect infestation from surrounding areas leading to thedecrease of the strength of the cross-section. This is especiallyproblematic for utility poles because they are typically canti-levered posts with the largest bending moment reaction at or

334 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

close to the ground level. These two factors create a conditionwhere wooden utility poles have a finite service life.

The decay model developed by Wang et al. (2008) isadopted in this work. The model, based on a study ofAustralian wood samples, incorporates several physical pa-rameters that include wood specie, chemical treatment, woodtype (sap/core/heartwood), and climate parameters (meanannual temperature and rainfall). The decay rate, run’ stake, foran untreated test stake is defined as:

run’stake¼ kwood kclimate ð1Þwhere kwood is a function of the durability class of the speciesand varies across the section of the wood. kclimate is a functionof the annual rainfall and mean temperature. A detaileddescription of the decay model is provided in Wang et al.(2008).

2.2. Climate data

Climate has a profound effect on the decay rate of wood.As seen in the decay model parameters, two climatic factorsare key to determining a time-based decay model, averagetemperature, and annual rainfall. The climate model data usedhere comes from the Coupled Model Intercomparison ProjectPhase 5 (CMIP5) multi-model ensemble (Maurer et al., 2007).CMIP5 provides precipitation and temperature data statisti-cally downscaled to finer resolutions using the monthly bias-correction and spatial disaggregation (BCSD) method. TwoRepresentative Concentration Pathways (RCP) are chosen tostudy the potential impact of climate change on decay andhurricane hazardRCP4.5RCP8.5. RCP4.5 is chosen to repre-sent a medium emission scenario, while RCP8.5 represents ahigh emission scenario. Seventy-one and 70 outputs fromdifferent Global Climate Models (GCMs) are available forRCP4.5 and RCP8.5, respectively. All the outputs in1950e2099 are used to account for uncertainties in themodels, as discussed in subsequent sections.

2.3. Pole failure risk

Wood poles supporting power lines can fail due to twocauses: hurricane wind (Type 1 failure) and excessive decay(Type 2 failure). Failure caused by hurricane wind depends onthe intensity of the wind load as well as the capacity of thepole, which is affected by time-dependent deterioration. Polesthat fail due to hurricanes cannot be repaired using FRP andare, therefore, replaced with new poles. Type 2 failure occurswhen the strength of the pole falls below a threshold due toexcessive decay. The threshold considered here is the NESCrecommended threshold, as discussed earlier, which is thepoint at which the strength of a pole falls below two-third ofthe initial strength. The methods to evaluate the two failuremodes are discussed below. Poles that undergo Type 2 failureare repaired using the FRP method described above.

2.3.1. Type 1: Failure due to hurricane windThe probability of failure of a pole due to hurricane wind is

computed by convolving the fragility of the poles and thehazard probability, as shown in Eq. (2).

Pf1ðtÞ¼Z

FDðV ; tÞ fvðV; tÞ dV ð2Þ

where Pf1ðtÞ is the probability of pole failure due to hurricanewind, FDðV; tÞ is the time-dependent fragility of the poles, andfvðV ; tÞ is the time-dependent hazard function. Hence, theimpact of wood decay and hurricane hazard are considered incalculating Pf1ðtÞ. A method for the time-dependent fragilityanalysis of poles has been developed and is detailed in Salmanand Li (2016a).

Because of the unpredictability of hurricane wind speedsthroughout the service life of a pole, wind must be modeled asa random variable with a time-sensitive probability distribu-tion. As such, hurricane simulation is typically employed tomodel hurricane hazard. The hurricane simulation modeldeveloped by Xu and Brown (2008) is adapted. The requiredparameters for the hurricane simulation in Florida and SouthCarolina are taken from Huang et al. (2001) and Xu andBrown (2008). The parameters for New York City are foundby fitting probability distributions to histograms of the pa-rameters from Lin et al. (2010). A summary of the simulationparameters is given in Table A1. The radial wind field modeldeveloped by Holland (1980) is used to calculate the gradientwind speed at locations of interest. The gradient wind speed isconverted to a surface and then 3-s gust wind speed usingfactors of 0.8 and 1.287, respectively (Vickery et al., 2009; Xuand Brown, 2008). The rise in central pressure of the hurricaneafter landfall is modeled using the approach developed byVickery and Twisdale (1995). Wind speed decay after landfallis accounted for using the models developed by Kaplan andDemaria (1995) and Kaplan and Demaria (2001). The radiusto maximum wind is modeled using the equation from FEMA(2011). The simulation is carried out for 200,000 hurricaneseasons. More details of the simulation procedure can befound in Salman and Li (2016b).

At the moment, climate models seem inconsistent on theprojection of changes to hurricane frequency due to climatechange (Knutson et al., 2010; Landsea et al., 2010), but mostmodels project an increase in intensity (Emanuel, 2005;Knutson et al., 2008; Oouchi et al., 2006; Staid et al., 2014;Stewart et al., 2014; Webster et al., 2005). While the increasein global intensity is projected to be between 2% and 11%(Knutson et al., 2010) under the status-quo assumptions,changes in the global intensity and frequency may be signifi-cantly different from the changes observed in individualtropical cyclone basins. For this study, only the northernAtlantic basin is considered as this is the area that feeds stormsaffecting eastern U.S. Webster et al. (2005) have shown thatfor most areas, the annual frequency and duration of hurri-canes is not statistically significant from zero, while the North

335MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Atlantic basin shows an increase in frequency with a 99%significance confidence level. However, the majority of liter-ature concerning hurricane frequency with respect to climatechange is still inconclusive, as summarized by Knutson et al.(2010). This is because the relation between anthropogenicforcing and hurricane activities is not well understood. Hence,there is no unified method of modeling climate change impacton hurricanes, and there is no consensus in the publishedliterature on even the direction of the change. Based on a re-view of the literature, the range of change in frequency andintensity in the Atlantic Basin and the U.S. east coast underRCP8.5 are �25% to þ55% and 0e25%, respectively. Thecorresponding values for frequency and intensity for RCP4.5are �62% to þ58% and �10% to þ12%, respectively.

From the studies mentioned above, this work assumes thatthe rise of intensity of hurricanes for the eastern U.S. willincrease linearly by 10% and 20% by the end of the 21stcentury for the RCP4.5 and RCP8.5 scenarios, respectively(Mudd et al., 2014b). Due to the large amount of discrepancyfor the projected increase or decrease in frequency, no changeis assumed between now and 2099.

2.3.2. Type 2: Failure due to excessive decayType 2 failure is evaluated as the probability that the

strength of a pole at the time of inspection is less than two-thirds of the initial strength as given by Eq. (3). Here, thestrength of the pole at the time of inspection is evaluated usingthe decay model discussed earlier. The strength at any time, t,is then a function of the decay rate, which in turn is a functionof climatic parameters (temperature and precipitation). Thetemperature and precipitation in any given year are modeled asnormally distributed random variables with mean and standarddeviation determined from all the outputs of the CMIP5 globalclimate models.

Pf2ðtÞ ¼ P

�MnðtÞ <

2

3Mn0

�ð3Þ

where Pf2ðtÞ is the probability that pole strength is less thantwo-thirds of the initial strength at any time t, MnðtÞ is thestrength of the pole at time t, and Mn0 is the initial strength ofthe pole obtained from ANSI-O5.1 (ANSI, 2002).

The NESC requires utility companies to regularly inspectand maintain the poles supporting power lines (IEEE, 2017).To comply with the NESC requirement, utility companiestypically carry out periodic inspections to identify poles thatneed maintenance. Considering the vast number of poles indistribution systems, it is neither practical nor economicallyfeasible to inspect all the poles in a system frequently. Utilitycompanies typically inspect a small percentage of the poles intheir network every year. The number of years it takes to finishinspecting all the poles and start over is the length of the in-spection cycle. For example, Florida Power and Light inspects1/8 of its pole population every year as part of an eight-yearpole inspection cycle plan. In a survey of 261 North Amer-ican utilities, Mankowski et al. (2002) reported that theaverage inspection cycle for distribution poles is about eight

years. In this study, it is assumed that the poles are inspectedon an 8-year cycle.

2.4. FRP repair planning

The NESC recommends the replacement of wood polesonce they reach two-thirds of their initial strength due todecay. Rather than replacing the poles outright, it is possible toreinforce the base of the poles with FRP to restore some of thelost strength. The repair is beneficial if the cost of applying therepair can overcome the costs of replacing the poles whenadjusted for labor, materials, and interruption to networkperformance.

Saafi and Asa (2010) developed a remediation techniqueusing a wet lay-up FRP sleeve applied at the ground lineextending three-pole diameters in length and tested onsouthern pine class 4 poles. Intact poles of the same age wereused to compare the effects of the FRP repair. After theapplication of the FRP to poles below the 67% initial strengththreshold, 85% of the initial strength had been restored. Asmentioned by Saafi and Asa (2010) the FRP jacket may reducethe rate of decay from the outer layers of wood by limiting themoisture and oxygen coming into contact with the pole. WhileSaafi and Asa (2010) assumed that the strength of the FRPjacket will deteriorate with time, there is no data on the extentto which the FRP will be weakened. Hence, it is assumed herethe FRP strength does not deteriorate with time. If informationon the deterioration of the FRP becomes available, it caneasily be incorporated in the time-dependent reliability of thepoles. There is, however, strong evidence that the repair ofstructural components using FRP can significantly decreasethe strength deterioration of the component (Debaiky et al.,2002; Masoud and Soudki, 2006; Pantazopoulou et al.,2001). Hence, it is assumed in this study that the decay rateof the pole will be halved after the application of the FRP. Asensitivity analysis is carried out to determine the effects ofthis assumption on the results, which will be discussed insection 3.6.

The FRP wrap had been determined by Saafi and Asa(2010) based on experimental data to increase the capacityby about 30% of the initial strength of the pole. To accountfor uncertainty in the strength added by the FRP repair, auniform distribution is assumed with the added strengthranging from 20% to 40% of the initial strength. It isacknowledged that the level of improvement in strength dueto repair by FRP will depend on factors such as the type ofFRP used and how it is installed, among others. Hence, weassumed that the FRP and installation method from Saafi andAsa (2010) are used.

The planning of the FRP repair can be carried out in twoways. The first approach is the threshold-based maintenanceplanning approach in which the FRP is applied, when thedecay of the pole reaches a certain threshold. In suchapproach, the empirical decay model discussed earlier can beused to determine the time at which the strength will fallbelow a certain threshold (the NESC threshold for example).The FRP repair can then be planned. In the absence of a decay

336 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

model, periodic inspection can be carried out to determinewhether the threshold for repair is reached.

The second approach for the repair planning is throughoptimization. In this approach, the repair is planned to opti-mize a specific parameter over the lifespan of the pole. In sucha case, the time of the repair will depend on the objective ofthe optimization. For example, the objective of the optimiza-tion can be to maximize the structural reliability of the pole,i.e., minimize the probability of failure. Another objective,which is more common in maintenance planning for structuresand infrastructure, is to minimize the maintenance cost overthe lifecycle or within a specific timeframe. A multi-objectiveoptimization can also be considered. In such a case, trade-offamong the various objectives is required. As this study isfocused on the structural performance of the poles consideringdecay over time, the FRP repair planning here will be carriedout through optimization considering structural reliability overa specific period as the objective. The optimization problemover a planning period T can be formulated as:

Given : Pole configuration; decay model; PH ¼ TFind : t ¼ tFRPTo minimize : Pf1ðT; tFRPÞ or Pf2ðT ; tFRPÞSubject to : Integer tFRP in ½0; TÞ yearsWhere PH is the planning horizon, tFRP is the time of the FRPrepair, Pf1 is the probability of Type 1 failure, Pf2 is theprobability of Type 2 failure. Recall that Type 1 failure con-siders both decay and hurricane hazard while Type 2 failure isbased on only decay. In locations with hurricane risk, Type 1failure can be used for the optimization. In locations with nohurricane risk, Type 2 failure can be used. The optimizationproblem here is complex and cannot be expressed by a closed-form expression containing the decision variable. Hence, theoptimization problem can be solved using either numericalexhaustive search method or heuristic algorithms such asparticle swarm optimization and genetic algorithm. In thiscase, the decision variable (optimal repair time or year) isdiscrete and the solution space is bounded within ½0; TÞ (e.g.,2010 to 2100). Therefore, a numerical exhaustive search isfeasible for the current problem. A MATLAB algorithm isdeveloped to solve the optimization problem exhaustively.

3. Case study and results

3.1. Pole description

The most common species of wood used for utility distri-bution is southern pine (Mankowski et al., 2002). As such, asouthern pine pole is considered. The decay rate describedearlier is a function of the pole durability class, which is basedon the average expected service life of the pole. Treated polessampled in Florida have shown evidence of poles working upto 50 years (KEMA, 2006). Hence, in this study, a durabilityclass 1 is assumed based on the classification employed in thedecay model adapted (Wang et al., 2008). Similarly, despitesouthern pine being defined as a softwood, the decay rate of

sapwood behaves as if it were a hardwood. This study assumesthat the FRP sleeve is meant to be applied to treated poles inthe field, with the treatment applied to the sapwood. The timelag for decay to begin has a high degree of uncertainty, so forthis study, the average decay rate across all three locations forheartwood was used to estimate the number of years beforedecay begins and rounded to the nearest year. The estimatedtime lag is approximately five years. This approach isconsidered as the effect of the time lag on the overall lifeextension of a pole is insignificant and the decay rate for alllocations is very small in the early years.

ANSI-O5.1 (ANSI, 2002) gives a minimum circumferencefor poles to be used for each class. A Class 4 southern pinepole is chosen, which has a minimum circumference of63.5e105.4 cm depending on the length of the pole of6.1e21.3 m. For the purpose of calculating the moment ca-pacity at the ground line, where the maximum stress occurs, apole length of 13.7 m is used, which is typical for distributionlines. This yields a minimum circumference at 1.8 m from thebutt of the pole to be 89 cm or a diameter of 28.3 cm. The poleis assumed to support 3 ACSR conductors with a diameter of18.3 mm located 0.3 m from the tip. It also supports one AACneutral with a diameter of 11.8 mm at 1.5 m from the tip. Thespan between the poles for wind pressure calculation isassumed to be 46 m (Short, 2005).

Sapwood decays at a different rate compared to heartwood.The thickness of sapwood in southern pine varies and has ahigher prevalence in timber with a faster growth rate. Thethickness of the sapwood can be described using the diameterof the outside bark using Eq. (4) (Schultz, 1997).

Tsap ¼ 0:36dbark � 0:224 ð4Þ

where dbark is the outside bark diameter at breast height (incm) in 27-year-old southern yellow pine trees as demonstratedby Schultz (1997). The assumed exterior diameter is based onthe ANSI-O5.1 (ANSI, 2002) classification of class 4 poles.For the analysis, dbark is approximated to be 27 cm based onstatistical data of loblolly pine (Schultz, 1997) and height ofclass 4 poles. This data is then assumed to be similar for otherspecies of southern yellow pine trees, which are commonlyused in the U.S. for wood construction and electrical poles.

The decay model described in Wang et al. (2008) mentionsvarious methods of decay direction (external vs. internal).Decay may progress from both the center progressing out-wards or from the pole surface and progress inward. For thisstudy, the decay pattern is only modelled progressing from thesurface inward.

3.2. Case study locations

Three locations have been selected to analyze the impact ofclimate change on hurricanes and the decay rate, and thus thelifecycle, of wood poles. Miami, Florida, Charleston, SouthCarolina, and New York City, New York, are selected to give awide breadth of climates. Miami has a tropical monsoonclimate, which lends itself to being hot and humid year-round.

337MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Because of its location between the Gulf of Mexico and theAtlantic Ocean, it is the most likely city in the U.S. to beaffected by tropical storms and hurricanes. Charleston has amilder subtropical climate with hot summers and mild winters.Charleston still has above average humidity and is prone totropical cyclones that follow the east coast. However, histor-ically, there have been no Category 5 storms to hit Charleston,but a few Category 4 events have occurred. New York City hasa much colder continental climate meaning it has hot summersand cold winters (Peel et al., 2007). The hurricane risk in NewYork City is lower than that of Charleston. The average pro-jected temperature and precipitation from the CMIP5 for thethree locations are shown in Fig. A1. Using the projected data,the mean decay rate of heartwood and sapwood are calculatedand shown in Fig. 1. As temperature and precipitation increasedue to climate change, so does the decay rate of the poles. It isnoted that the decay rate is affected by local environmentalconditions where the higher temperature and precipitation inMiami accelerate the deterioration process. To illustrate theeffects of climate change, Climate data from 1950 to 1970 isused for the no climate change scenario plot.

3.3. Impact of FRP repair time on strength

The period of interest is 2010e2099. FRP repair can becarried out at any time while the pole is in service. One repairscenario will be applying the FRP wrap when the pole'sstrength reaches two-third of its initial value, which is theNESC threshold for repair or replacement. Fig. 2 shows a plotof the variation of pole strength with time. The time it takes forthe strength to reach the NESC threshold is defined as theservice life of the pole. For illustration purposes, the addedstrength due to FRP repair is assumed to be 30% of the initialpole strength. From the analysis, the average service life of thepole in Miami is around 27 years with no climate change ef-fect. The average service life in Miami for both RCP4.5 andRCP8.5 is the same as the no climate change scenario. This isbecause the difference between the two in terms of changes indecay rate only becomes apparent later in the 21st century, asseen in Fig. 1. Even towards the end of the 21st century, theincrease in decay rate due to climate change does not have asignificant impact on the change in diameter of the poles andthe strength, as seen in Fig. 2. For example, the service life ofthe pole in Miami only decreased from 62 years for the noclimate change scenario to 59 years for the RCP8.5 scenario.

The application of the FRP adds about 35 years to theservice life of the pole in Miami with no climate change and33 and 32 years for RCP4.5 and RCP8.5, respectively. TheFRP adds about 49, 44, and 42 years to the service life inCharleston with no climate change, RCP4.5, and RCP8.5,respectively. The corresponding additions are greater than 52,51, and 49 years in New York City. The variation in the servicelife in the three locations is due to the varying climates in theregions with poles in Miami having the highest decay rate dueto its tropical monsoon climate.

Note that the sudden change in the slope of the lines afterabout 70 years is due to the transition from the decay of the

sapwood to the heartwood. Because the kwood factor in Eq. (1)is lower for the heartwood, the transition leads to a decrease indecay rate, which resulted in a decrease in strength loss seen inFig. 2.

As mentioned previously, the CMIP5 provides 71 and 70outputs from different Global Climate Models (GCMs) forRCP4.5 and RCP8.5, respectively. The results shown in Fig. 2are based on the ensemble means of the 71 and 70 outputs. InFig. 3, the change in strength with time for the individualGCM outputs is plotted. The plots in Fig. 3 give an indicationof the model uncertainty from the GCMs. The model uncer-tainty is associated with the fact that different GCMs providedifferent outputs for a given radiative forcing depending on thesimplifications and assumptions used in the models (Bastidas-Arteaga and Stewart, 2015). It can be seen from Fig. 3 that thevariation in the results considering all the GCM outputs issmall. In subsequent analyses, the temperature and precipita-tion are modeled in a Monte Carlo simulation as normallydistributed random variables with mean and standard deviationdetermined at each time from all the GCM outputs, asmentioned earlier.

The FRP can also be installed at any time before thestrength of the pole reaches the NESC threshold. In Fig. 4,the time of repair is varied along the pole's service life todetermine if there is an optimal time of repair. The figureshows the relation between the service life of the pole andFRP repair age. For Fig. 4, the added strength by the FRP isassumed to be 30% of the initial strength of the pole forillustration purposes. From the results, it is apparent thatthere is a positive correlation between earlier repair time andextended pole life. This is most likely due to the assumptionthat the presence of the FRP reduces the amount of moistureand fungal contact with the underlying pole, which producesa slower decay rate than merely treating the poles. The onlyexception to this trend is for Miami, in which there is anuptick of service life for FRP installation at the NESCthreshold (27 years) compared to installation at 20 years.This is most likely due to the projections that precipitation isnot expected to drastically increase in the RCP4.5 andRCP8.5 conditions in Miami with the RCP8.5 scenarioactually projecting that precipitation will decrease (Fig. A1).In all cases, the service life could not be accurately estimatedpast the 90-year mark as the CMIP5 data only extendsthrough the end of the 21st century. Hence, in cases where theservice life extends beyond the end of the 21st century afterFRP repair, the service life is not estimated. Consequently,only two points of the New York City results are plotted. Theservice life of the poles for repair at 10 and 20 years extendsbeyond the 90-year mark.

3.4. Type 1 failure risk results

Figure 5 illustrates the change in the probability of Type 1failure (due to hurricane winds) for each of the three zonesstudied for three repair scenarios: no FRP, FRP installation at16 years, and FRP installation at 32 years. To determine thefailure probabilities here and in subsequent sections, the gain

Fig. 1. Mean decay rate, left: sapwood, right: heartwood in 1950e2099 (No CC is no climate change scenario).

338 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Fig. 2. Average strength plot with FRP repair at NESC strength threshold (No

CC is no climate change scenario).

339MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

in strength due to FRP repair is assumed as random asdescribed in Section 2.4. The figure illustrates that the effec-tiveness of FRP repair is not only time-dependent but locationor climate-dependent as well. This is most readily apparent inthe New York City cases in which the difference betweenapplying the FRP coat at 16 or 32 years do not differ greatly.However, the decision to apply the FRP sleeve reduces theprobability of failure in New York City from about 4% in 2099

in the no-FRP and no climate change scenario to less than 1%for both scenarios in which FRP sleeve is used as evidenced inFig. 5.

The results in Fig. 5 demonstrate that climate changedramatically increases the probability of failure, subsequentlylowering the lifespan of wooden poles. RCP4.5 and RCP8.5scenarios increased the failure probabilities in Miami at theend of the 21st century by about 15% and 30%, respectively,compared to the no climate change scenario. The corre-sponding increases are 23% and 70% in Charleston, and 25%and 73% in New York City. The comparatively lower increasein Miami is due to a decrease in precipitation towards the endof the 21st century for the RCP8.5 scenario and little increasein precipitation for the RCP4.5 scenario, as seen in Fig. A1.

While the decision to implement the FRP repair alwaysimproves the lifespan of wooden utility poles, as seen inFig. 5, the optimal time of repair is dependent on the location.In most cases, there was a positive relationship between earlyrepair and extension of effective service life. The exception tothis being applying FRP at the time of replacement for Miami.This again is most likely due to the predicted decrease inprecipitation in Miami. It is seen that the application of theFRP sleeve is effective in mitigating the probability of failureof the poles due to the RCP4.5 and RCP8.5 scenarios.

Note that the parameters used in the time-dependentfragility analysis to obtain the annual probabilities of failureare summarized in Table A2.

3.5. Type 2 failure risk results

As mentioned in Section 2.4.2, Type 2 failure is defined bypassive decay eroding the available strength of electrical polesto an unacceptable state as defined by the NESC. Fig. 6 il-lustrates the probability of Type 2 failure for each locationover the service life of the pole. The analysis assumes that allpoles were installed, and subsequently began their decayprocess, in 2010 and modeled using climate parameters basedon the CMIP5 scenarios mentioned previously through the endof 2099.

It can be seen from the left column in Fig. 6 that climatechange has little impact on Type 2 failure probability. This isbecause the increase in decay rate due to climate change is notlarge enough to have a significant impact on the decrease indiameter and, subsequently, the strength of the pole. This isevident from Fig. 2, where it is seen that climate change haslittle impact on the strength of the pole over time. Forexample, the pole used in the case study has a diameter ofabout 283 mm. The average decay rate of the sapwood inMiami increased gradually from about 1.83 mm per year in2010 to about 1.94 mm per year in 2099 under RCP8.5, asseen from Fig. 1. The corresponding increase for the heart-wood is from about 0.65 to about 0.7 mm per year. Thisresulted in a negligible impact on the failure probabilities ofthe pole. Note that Figs. 1 and 2 are plotted using averageannual temperature and precipitation, while in Figs. 5 and 6,the temperature and precipitation are sampled from a normal

Fig. 3. Strength plots of all CMIP5 outputs with no FRP repair.

340 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Fig. 4. Estimated average service life for different repair times (No CC is no climate change scenario).

Fig. 5. Type 1 failure probabilities with no FRP (left), FRP installation at 16 years (middle), and installation at 32 years (right) (No CC is no climate change

scenario).

341MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Fig. 6. Type 2 failure probabilities with no FRP (left) and with FRP installed at 24 years (right) (No CC is no climate change scenario).

342 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

distribution each year based on the outputs from variousGCMs in the CMIP5 records.

Comparing the results with no FRP and with FRP, the caseswith FRP repair show a higher impact of climate change onfailure probability (comparing the two figures for Charleston,

for example). This is because the application of the FRPcauses decay towards the end of the 21st century to occur inthe sapwood, which has a higher decay rate compared to theheartwood. In the case with no FRP, however, the sapwoodwill decay completely by the middle of the century, and the

343MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

decay will attack the heartwood, which has a much lowerdecay rate, as seen in Fig. 1. This transition will significantlydecrease the decay rate despite climate change resulting inlittle impact of climate change on the failure probability by theend of the 21st century.

When comparing the results from Fig. 6 to that of Fig. 5,the largest discrepancy is that the spread of the results in Fig. 6is significantly lower. The reason is that for the Type 1 failureprobability, the limit state function is a convolution of thefailure probability given a wind speed with the probability ofthe wind speed occurring. The combined impact of climatechange on both decay and hurricane hazard in Fig. 5 resultedin a much higher impact on Type 1 failure probabilities. Thisis further discussed in the subsequent sensitivity analysissection.

Note that the initial strength of the class 4 southern pinepole selected for the case study is 55.2 MPa and is lognormallydistributed with a coefficient of variation of 20%, according toANSI-O5.1 (ANSI, 2002). Because of the large coefficient ofvariation, the probability that the strength of a new pole is lessthan two-third of 55.2 MPa is about 2%. That is why thefailure probabilities in Fig. 6 for new poles is about 2%.

3.6. Sensitivity analysis

As mentioned previously, studies have been done to predictthe magnitude of changes to tropical cyclone intensity andfrequency through the end of the 21st century (Emanuel, 2005;Knutson et al., 2008; Mudd et al., 2014b; Oouchi et al., 2006;Webster et al., 2005). Based on these results, a linear increasein intensity of 10% by the year 2099 for the RCP4.5 scenarioand 20% for the RCP8.5 scenario with no change in frequencywas assumed. To investigate the sensitivity of Type 1 failure(due to hurricane wind) to only changes in decay rate due toclimate change, the failure probabilities are evaluatedassuming that there is no increase in the intensity of hurri-canes. Only the impact of climate change on decay rate isconsidered. The results are shown in Fig. 7. Comparing Fig. 7to Fig. 5 exemplifies the impact that an increase in storm in-tensity has on the failure probabilities. When only taking intoaccount the impact of climate change on decay for the twoRCP scenarios (Fig. 7), the increase in the probability offailure without FRP by the end of 2099 under RCP8.5 scenariorange from about 5% in Miami to about 22% and 20% inCharleston and New York City compared to a no climatechange scenario. Adding increases in the probability of moreintense wind events (Fig. 5), however, yields an increase infailure probabilities of the baseline no-FRP case of about 30%,70%, and 73% in Miami, Charleston, and New York City,respectively.

As previously mentioned, Saafi and Asa (2010) assumedthat after the application of an FRP sleeve, the rate of decay ofthe poles is roughly half of the decay rate prior to the instal-lation of the sleeve. To investigate the impact of the

assumption on the results, a sensitivity analysis is carried outassuming the FRP has no effect on the decay rate of the poles.The results are shown in Fig. 8 for Type 2 failure. Theassumption has a significant impact on the results. In NewYork City for example, the probabilities increased from around0.55 to more than 0.9. The results illustrate an additionaladvantage of the FRP repair. Not only does it restore thestrength of the poles, but it can slow down the decay rate and,consequently, extend the service life of the poles. Note thatwhile the extent to which the FRP slows down the decay isunknown in this case, there is significant evidence that therepair of structural components using FRP can significantlydecrease the strength deterioration of the component asmentioned previously (Debaiky et al., 2002; Masoud andSoudki, 2006; Pantazopoulou et al., 2001).

3.7. FRP repair optimization results

The objective of the optimization is to find the FRP repairtime that will minimize the failure probability over theconsidered period (2010e2099). Fig. 9 shows the FRP repairoptimization results, where the maximum failure probabilityvariation over the considered period for different repair timesis plotted. Note that only Type 1 failure probability isconsidered because the case study locations have hurricanerisk. Recall that Type 1 failure includes the effects of bothwood decay and hurricane hazard. For locations without hur-ricane hazard, the optimization can be carried out in a similarmanner using Type 2 failure probability. The plots on the leftside of Fig. 9 show the results based on the previously statedassumption that the application of the FRP will slow down thewood's decay rate by half. The results show that the optimalFRP installation time is at the beginning of the pole's lifecycle, i.e., at the end of the decay lag period. This is becauseof the assumption of the significant decrease in the decay rateof the wood after FRP installation. Early FRP installationleads to slow decay throughout the considered period, whichsignificantly decreases the failure probability. The plots inFig. 5 confirm this trend.

The results on the right side of Fig. 9 show the optimalrepair time if the FRP installation has no effect on the decayrate of the wood. In this case, the optimal repair time is not atthe beginning of the pole's life cycle. Rather, it is more thanhalfway through the life cycle in all cases. The results showthat it is better to allow the pole to decay significantly beforethe repair is carried out. The optimal repair times in Miami,Charleston, and New York City with no climate change are 58,59, and 59 years, respectively. While climate change has animpact on the failure probabilities, as seen earlier, it does notaffect the optimal FRP repair time. In Miami, the optimalrepair times are 58, 57, and 59 years, for no climate change,RCP4.5, and RCP8.5, respectively. The corresponding repairtimes are 59, 58, and 60 years in Charleston; and 59, 59, and60 years in New York City.

Fig. 7. Type 1 failure probabilities with no FRP (left), FRP installation at 16 years (middle), and installation at 32 years (right) without hurricane intensity increase (No CC is no climate change scenario).

344

MERSCHMAN

E.et

al./Advances

inClim

ate

ChangeResea

rch11(2020)332e

348

Fig. 8. Type 2 failure probabilities with FRP installed at 24 years, assuming a decrease in wood decay rate after FRP repair (left), assuming no change in wood

decay rate after FRP repair (right) (No CC is no climate change scenario).

345MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

Even though the decay rates in the three considered lo-cations vary, the results on the right of Fig. 9 show that theoptimal repair time is independent of location. This isbecause the optimization is based on the maximum failureprobability over the considered period. Hence, the optimalrepair year depends on the point at which the failure

probability at the end of the 21st century and the year of therepair will be minimized. In other words, the optimization isabout finding a balance point to keep the failure probabilityfrom becoming too large over the 90-year period. Hence, themagnitude of the failure probability does not influence theoptimal repair time.

Fig. 9. FRP repair optimization results for type 1 failure probabilities, assuming a decrease in wood decay rate after FRP repair (left), and assuming no change in

wood decay rate after FRP repair (right) (No CC is no climate change scenario).

346 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

4. Conclusions

This study investigates the effectiveness of using fiber-reinforced polymer (FRP) sleeve to reinforce wood polessubjected to decay to restore their lost strength and extend

their useful service life. The potential impact of climatechange on the pole decay rate and the intensity and frequencyof hurricanes is also considered. Additionally, the optimal FRPrepair time based on structural reliability consideration isdetermined. Poles located in Miami, Charleston, and New

347MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

York City, are considered in a case study. The results of thecase study show that FRP is highly effective in extending thepole's service life. Repair using FRP extended the service lifein Miami, Charleston, and New York City by about 35, 49, and>52 years. The higher increase in service life in Charlestonand New York City is because of lower pole decay ratesrelative to Miami. The results also show that the effect ofclimate change is highly related to the location under inves-tigation. In Miami, where precipitation was projected todecrease over time, the increase in failure probability at theend of the 21st century due to decay and hurricane hazardunder RCP4.5 and RCP8.5 emission scenarios was 15% and30%, respectively. In Charleston and New York City, however,where both temperature and precipitation were projected toincrease over time, the corresponding increases are 23% and70%, and 25% and 73%, respectively. It is also concludedfrom the results that while the incremental change in climateconditions does not significantly impact the rate of decay ofpoles, the increased intensity of storms over time does have asignificant impact on the structural reliability of the poles overtime. This conclusion is true for all three locations considered.

Based on the optimization results, it can be concluded thatthe optimal FRP repair time depends on the impact of the FRPon the decay rate of the wood. If the FRP can significantlydecrease the decay rate, the optimal repair time is at thebeginning of the pole's life cycle. However, if the FRP has noimpact on the decay rate, then it is better to repair the poleafter significant decay. Hence, more experimental data isrequired to characterize the kinetics of the decay rate afterrepair. In addition, the durability of the bonding between thetimber and the FRP as well as the durability of the FRP shouldbe considered in the future. It should also be noted that theoutcome of the optimization could change when cost isincluded.

In general, the results show that the application of an FRPreinforcement sleeve is effective in decreasing the risk due toboth deterioration and wind loading. While reinforcement ofwooden electrical distribution poles has been common prac-tice, typically, reinforcement is in the form of metal plates orsplints to shore up the ground-level cross-section. The majorbenefit of using an FRP solution is that it can simultaneouslyprotect and strengthen poles from multiple threats. While steelsheets may increase the cross-sectional moment capacity, theplates are typically not installed around the entire circumfer-ence of the pole, which still allows for fungal and moistureattacks to the section. Additionally, the effectiveness of steelplates has a high degree of uncertainty as it is difficult topredict how well the plates are attached or bonded to the poles.

It should be noted, however, that the results of the casestudy are based on a specific decay model, locations, andclimate change scenarios. The decay model adopted in thisstudy uses yearly averaged data. Consequently, the effects ofextreme daily or seasonal conditions that could significantlyincrease the decay rate are not captured. More precise pre-dictions will require a more advanced decay model able toaccount for daily or seasonal data. However, the methodproposed in this study is general and can be used with any data

for the sake of wood pole asset management by utilitycompanies.

Declaration of competing interest

The authors declare no conflict of interest.

Acknowledgments

We acknowledge the modeling groups, the Program forClimate Model Diagnosis and Intercomparison (PCMDI) andthe WCRP's Working Group on Coupled Modeling (WGCM)for their roles in making available the WCRP CMIP3 andCMIP5 multi-model dataset. Support of this dataset is pro-vided by the Office of Science, U.S. Department of Energy.

Appendix A. Supplementary data

Supplementary data to this article can be found online athttps://doi.org/10.1016/j.accre.2020.10.001.

References

ANSI (American National Standards Institute), 2002. Wood poles specifica-

tions and dimensions, vol. 1. ANSI-O5., Washington, DC.

Balsamo, A., Colombo, A., Manfredi, G., et al., 2005. Seismic behavior of a full-

scale RC frame repaired using CFRP laminates. Eng. Struct. 27 (5), 769e780.

Bastidas-Arteaga, E., Stewart, M.G., 2015. Damage risks and economic

assessment of climate adaptation strategies for design of new concrete

structures subject to chloride-induced corrosion. Struct. Saf. 52, 40e53.

Brischke, C., Rapp, A.O., 2008. Dose: response relationships between wood

moisture content, wood temperature and fungal decay determined for 23

European field test sites. Wood Sci. Technol. 42 (6), 507.

Chahrour, A.H., Soudki, K.A., 2006. Structural retrofitting of deteriorated

concrete lighting poles using FRP sheets in wet layup: field application. J.

Compos. Construct. 10 (3), 234e243.Cui, W., Caracoglia, L., 2016. Exploring hurricane wind speed along U.S.

Atlantic coast in warming climate and effects on predictions of structural

damage and intervention costs. Eng. Struct. 122, 209e225.

Debaiky, A.S., Green, M.F., Hope, B.B., 2002. Carbon fiber-reinforced poly-

mer wraps for corrosion control and rehabilitation of reinforced concrete

columns. Mater. J. 99 (2), 129e137.

Emanuel, K., 2005. Increasing destructiveness of tropical cyclones over the

past 30 years. Nature 436 (7051), 686.

Emanuel, K., Sundararajan, R., Williams, J., 2008. Hurricanes and global

warming: results from downscaling IPCC AR4 simulations. Bull. Amer.

Meteor. 89 (3), 347e368.FEMA (Federal Emergency Management Agency), 2011. Multi-hazard loss

estimation methodology, hurricane model: Hazus-MH 2.1 technical

manual. Washington, DC.

Ghobarah, A., Said, A., 2001. Seismic rehabilitation of beam-column joints

using FRP laminates. J. Earthq. Eng. 5 (1), 113e129.

Grace, N.F., Sayed, G., Soliman, A., et al., 1999. Strengthening reinforced

concrete beams using fiber reinforced polymer (FRP) laminates. ACI

Struct. J. 96 (5), 865e874.Gutowski, J., Hegerl, C., Holland, J., et al., 2008. Causes of observed changes

in extremes and projections of future changes. Weather and climate ex-

tremes in a changing climate. CCSP Synth. Assess. Product 3 (3), 81e116.

Holland, G.J., 1980. An analytic model of the wind and pressure profiles in

hurricanes. Mon. Weather Rev. 108 (8), 1212e1218.

Huang, Z., Rosowsky, D., Sparks, P., 2001. Hurricane simulation techniques

for the evaluation of wind-speeds and expected insurance losses. J. Wind

Eng. Ind. Aerod. 89 (7), 605e617.

348 MERSCHMAN E. et al. / Advances in Climate Change Research 11 (2020) 332e348

IEEE (Institute of Electrical and Electronics Engineers), 2017. National

electric safety code. New York, NY.

Kaplan, J., Demaria, M., 1995. A simple empirical model for predicting the

decay of tropical cyclone winds after landfall. J. Appl. Meteorol. 34 (11),

2499e2512.Kaplan, J., Demaria, M., 2001. On the decay of tropical cyclone winds after

landfall in the New England area. J. Appl. Meteorol. 40 (2), 280e286.

Kell, J.A., 2001. Repair of wooden utility poles using fibre-reinforced poly-

mers. University of Manitoba, Winnipeg.

KEMA, 2006. Post hurricane wilma engineering analysis. https://www.

coursehero.com/file/34562321/FPL-Pre-Workshop-Responses-Wilma-1doc/.

Knutson, T.R., Sirutis, J.J., Garner, S.T., et al., 2008. Simulated reduction in

Atlantic hurricane frequency under twenty-first-century warming condi-

tions. Nat. Geosci. 1 (6), 359.

Knutson, T.R., McBride, J.L., Chan, J., et al., 2010. Tropical cyclones and

climate change. Nat. Geosci. 3 (3), 157e163.Landsea, C.W., Vecchi, G.A., Bengtsson, L., et al., 2010. Impact of duration

thresholds on atlantic tropical cyclone counts. J. Clim. 23 (10), 2508e2519.

Lin, N., Emanuel, K., Smith, J., et al., 2010. Risk assessment of hurricane

storm surge for New York City. J. Geophys. Res. Atmos. 115 (D18).

Lopez-Anido, R., Michael, A.P., Sandford, T.C., et al., 2005. Repair of wood

piles using prefabricated fiber-reinforced polymer composite shells. J.

Perform. Constr. Facil. 19 (1), 78e87.Mankowski, M., Hansen, E., Morrell, J., 2002.Wood pole purchasing, inspection,

and maintenance: a survey of utility practices. For. Prod. J. 52 (11/12), 43.

Mann, M.E., Emanuel, K.A., 2006. Atlantic hurricane trends linked to climate

change. Eos Trans. AGU. 87 (24), 233e244.

Masoud, S., Soudki, K., 2006. Evaluation of corrosion activity in FRP repaired

RC beams. Cement Concr. Compos. 28 (10), 969e977.

Maurer, E.P., Brekke, L., Pruitt, T., et al., 2007. Fine-resolution climate pro-

jections enhance regional climate change impact studies. Eos Trans. AGU

88 (47). https://doi.org/10.1029/2007EO470006.

Mudd, L., Wang, Y., Letchford, C., et al., 2014a. Hurricane wind hazard

assessment for a rapidly warming climate scenario. J. Wind Eng. Ind.

Aerod. 133, 242e249. https://doi.org/10.1016/j.jweia.2014.07.005.

Mudd, L., Wang, Y., Letchford, C., et al., 2014b. Assessing climate change

impact on the U.S. East Coast hurricane hazard: temperature, frequency,

and track. Nat. Hazards Rev. 15 (3), 04014001.

Najm, H., Secaras, J., Balaguru, P., 2007. Compression tests of circular timber

column confined with carbon fibers using inorganic matrix. J. Mater. Civ.

Eng. 19 (2), 198e204.

Oouchi, K., Yoshimura, J., Yoshimura, H., et al., 2006. Tropical cyclone

climatology in a global-warming climate as simulated in a 20 km-mesh

global atmospheric model: frequency and wind intensity analyses. J.

Meteorol. Soc. Jpn. Ser. II 84 (2), 259e276.Pantazopoulou, S.J., Bonacci, J., Sheikh, S., et al., 2001. Repair of corrosion-

damaged columns with FRP wraps. J. Compos. Construct. 5 (1), 3e11.

Peel, M.C., Finlayson, B.L., McMahon, T.A., 2007. Updated world map of the

K€oppen-Geiger climate classification. Hydrol. Earth Syst. Sci. Discuss. 4

(2), 439e473.

Polyzois, D., Kell, J.A., 2007. Repair and rehabilitation of wood utility poles

with fibre-reinforced polymers. Can. J. Civ. Eng. 34 (1), 116e119.

Ranson, M., Kousky, C., Ruth, M., et al., 2014. Tropical and extratropical

cyclone damages under climate change. Climatic Change 127 (2),

227e241.

Ryan, P.C., Stewart, M.G., 2017. Cost-benefit analysis of climate change

adaptation for power pole networks. Climatic Change 143 (3e4),519e533.

Ryan, P.C., Stewart, M.G., Spencer, N., et al., 2014. Reliability assessment of

power pole infrastructure incorporating deterioration and network main-

tenance. Reliab. Eng. Syst. Saf. 132, 261e273.

Ryan, P.C., Stewart, M.G., Spencer, N., et al., 2016. Probabilistic analysis of

climate change impacts on timber power pole networks. Int. J. Electr.

Power Energy Syst. 78, 513e523.Saafi, M., Asa, E., 2010. Extending the service life of electric distribution and

transmission wooden poles using a wet layup FRP composite strengthening

system. J. Perform. Constr. Facil. 24 (4), 409e416.

Salman, A.M., Li, Y., 2016a. Age-dependent fragility and life-cycle cost

analysis of wood and steel power distribution poles subjected to hurri-

canes. Struct. Infrastruct. Eng. 12 (8), 890e903.

Salman, A.M., Li, Y., 2016b. Assessing climate change impact on system

reliability of power distribution systems subjected to hurricanes. J. Infra-

struct. Syst. https://doi.org/10.1061/(ASCE)IS.1943-555X.0000316.

Schultz, R.P., 1997. Loblolly Pine: the Ecology and Culture of the Loblolly

Pine (Pinus Taeda L.). U.S. Government Printing Office, Washington DC.

Sharif, A., Al-Sulaimani, G., Basunbul, I., et al., 1994. Strengthening of

initially loaded reinforced concrete beams using FRP plates. ACI Struct. J.

91 (2), 160e168.

Short, T.A., 2005. Electric Power Distribution Equipment and Systems. CRC

Press, Boca Raton.

Staid, A., Guikema, S.D., Nateghi, R., et al., 2014. Simulation of tropical

cyclone impacts to the US power system under climate change scenarios.

Climatic Change 127 (3e4), 535e546.

Stewart, M.G., Wang, X., Willgoose, G.R., 2014. Direct and indirect cost-and-

benefit assessment of climate adaptation strategies for housing for extreme

wind events in Queensland. Nat. Hazards Rev. 15 (4), 04014008.

Teodorescu, I., T‚�apus‚i, D., Erbas‚u, R., et al., 2017. Influence of the climatic

changes on wood structures behaviour. Energy Proc. 112, 450e459.

Triantafillou, T.C., 1998. Shear strengthening of reinforced concrete beams

using epoxy-bonded FRP composites. ACI Struct. J. 95, 107e115.Vickery, P.J., Twisdale, L.A., 1995. Wind-field and filling models for hurricane

wind-speed predictions. J. Struct. Eng. 121 (11), 1700e1709.

Vickery, P.J., Masters, F.J., Powell, M.D., et al., 2009. Hurricane hazard

modeling: the past, present, and future. J. Wind Eng. Ind. Aerod. 97 (7),

392e405.

Wang, C., Leicester, R., Nguyen, M., 2008. Manual No. 3: decay in ground

contact. Forest & Wood Products Australia.

Webster, P.J., Holland, G.J., Curry, J.A., et al., 2005. Changes in tropical

cyclone number, duration, and intensity in a warming environment. Sci-

ence 309 (5742), 1844e1846.

Xu, L., Brown, R.E., 2008. A hurricane simulation method for Florida utility

damage and risk assessment. Power and Energy Society General Meeting-

Conversion and Delivery of Electrical Energy in the 21st Century, 2008.

IEEE.